Linking metabolomics to machine learning reveals the metabolic fates of the refractory industrial pollutant 1-Hexadecene

In order to solve the low removal efficiency of 1-Hexadecene in industrial wastewater treatment plants and fill the knowledge gap of its microbial metabolic mechanisms, this work studied the 1-Hexadecene degrading microorganisms enrichment and its metabolic pathway using continuous bioreactors and batch incubation tests. With the successful enrichment of hexadecene-degrading microorganisms, the biodegradation rates of 1-Hexadecene at 50–200 mg/L could range from 58.6 % to 72.5 %. The 16S rRNA fragment and microbial community analysis identified Flavihumibacter, Gemmatimonas, and Caulobacter as putative 1-Hexadecene degraders. Through mass spectrum qualification analysis, Principal component analysis and orthogonal partial least squares discriminant analysis, totally 81 featured metabolites associated with 1-Hexadecene biogradation were screened out. The machine learning based support vector machines and random forests further identified 18 biomarkers during those featured metabolites. Based on the coupling metabolomics analysis with machine learning, three potential metabolic pathways with Heptane, 2,5-dimethyl hexane, 3-ethyl-3-methyl-heptane as the respective end products were proposed through biomarkers. These results provide new insights for the analysis of metabolic pathway of 1-Hexadecene, and theoretical support for the bioaugmentation of its treatment processes. © 2024 Elsevier B.V.